Document 3QN2jJ6j4RGN14oXMX9G4Lo3E

FROM SUBJECT Dr. J. B. PauschCorporate Technical Support - Brecksville Fiao POINT OR DEPT. & BLDG. NO. Research S Development Center- Brecksville PUBLICATION APPROVAL DATE THIS LETTER 1-24-83 / \ The BFGoodrich Publications Committee has approved for publication the Abstract of your proposed paper entitled "Pyrolysis Studies of Chlorinated Poly (vinyl chloride)." Our records indicate that the paper will be presented orally by Dr. Lattimer-at the 31st Annual Conference on Mass Spectrometry and Allied Topics in Boston, Massachusetts, May 8-13, 1983. And the paper will be published in Macro molecules. This is subject to approval of the fined, paper by the Publications Committee. ccy Research Files, Brecksville - w/Meuiuscript and approval sheets Public Relations, Akron -w/Manuscript E. G. Fiorito - w/Manuscript D. E. Ley - w/Copy of approval sheets tooeac.02 BFG-4956-E 11/80 LITHO IN U.S.A. BFG05143 .- - .* o:: 7' .v - ? ' r Messrs- K. C. Baranwal J. W. Messerly R. "ij. ` Meyer. - 2 4- C. E. Wilkes D. E. Ley - For SI Metric Conversion This paper may be of interest to your division. If so, would you comment on the suitability for publication on the attached Request for Approval sheet and return to me. wwrTT Tfi. ^ji:. "''Vj'***''z.-m'f-'z If this paper is not relevant to your division, please return indicating no interest. crvV-71 Promptness in reply will be appreciated. Attachment 7^_ /-3 ^ <iv < UX^~4\ " -:: - ... :--zyz: Ijnjr3 * r -. - 3:;-. BFG05144 . ^\ ' REQUEST FOR APPROVAL--F--O--R----P--U--B--L--IC---A--T-I-O---N- Title of Paper Pyrolysis Studies oflfLightly/Chlorinated Poly(vinyl chloride) Author(s)_____ R. P. Lattimer, J. B. 'gaiisch-*ind H. L. C. Meuzelaar Date January 5. 1983 Deadllne^Remarks 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolvsis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference is the best forum for presentation of new results in this field. 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in Pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published.____________________________________________________________________________ ______________ 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057;_____________ _____________ R. P. Lattimer. IOC, 10/14/82.____________________________________________________ / 6. In what ways is BFGoodrich using this information? The techniques are used to characterize complex polymer systems, 20729003 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation bv Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. May 8-13. 1983.__ Publication in Macromolecules ._____________ 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer.__________________________________ pvceived 'JAN 1319 R.J. FAWCETT w\^ W-.L C. SSiiggnnaattuure of Author(s) U Signature of Author's Manager/ BFG05145 Title of Paper Pyrolysis Studies of Chlorinated Polvfvinvl chloride) Author(s)R. P. Lattimer. J. B. Pausch and H. L. C. Meuzelaar Date February 2& 1983 Deadline Remarks to be presented May 8-13. 1983 Abstract approved January, 1983. 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference fs the best forum for presentation of new results in this field.;_________________ __ |, 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BF6 expertise in the development of new techniques for polymer characterization. It will show BF6 collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published.__________________________! 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R. P, lattimer. TOC. 10/14/82. R. P. lattimer and J. B. Pausch, Research Report. 1/17/83. 6. In what ways is BFGoodrich using this information? The techniques are, used to characterize complex polymer systems. 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. Mav 8-13. 1983. Publication in Macromolecules.__________ ________________ 1' t < - w * ! -v 9. Was this paper requested or volunteered? Volunteered by Lattimer.________ ______ \X` \ By Whom? MAR.3 _ !S33 H. J. v.-vwwgTT. jH-C^/Uxt^c f ,\ \Wy' Signature of Author(s) ! U r- 20729004 Signature of Author's Manager^/^*"" BFG05146 ). Title of Paper Pyrolysis Studies of Lightly Chlorinated Poly (vinyl chloride)______________ Author(s)R. P. Lattimer, J. B. Pausch and H. L, C. Meuzelaar____________________________ DateJanuary 5. 1983 Deadline Remarks _________________________________________________________ __ 1. Why do you feel this paper is a worthwhile contribution to the technical literature?^ It correlates analytical pvrolvsis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pvrolvsis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? // The annual mass spectrometry conference is the best forum for presentation of new results in this field. 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new-techniques for polymer characterization. It will show BFG collaboration with a recognized authority in Pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general'analytical techniques used have also been published. 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R. P. Lattimer, IOC, 10/14/82. 6. In what ways is BFGoodrich using this information? The techniques are used to characterize complex polymer systems. 20729005 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Tonics. Boston. May 8-13. 1983. Publication in Macromolecules 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer._________________________________ VVW JAN1J1983 R.J.FAWCETT . /1 Signature of Author(s) .O' <-- g-. ec .. rx^ '-i Signature of Author's Manage,!" 7 qXclCLA U prr BFG05147 i / Title of Paper Pyrolysis Studies of Lightly Chlorinated Poly (vinyl chloride) Author(s)______R. P. Lattimer,' J. B. Pausch and H. L. C. Meuzelaar DateJanuary 5, 1983 Deadline Remarks __________________________________________________________ 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pvrolvsis-mass spectrometry in combination with computerized Pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? // The annual mass spectrometry conference is the best forum for presentation of new results in this field. 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in Pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published._________________________________________________________________________________________ 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057;__________ R. P. Lattimer. IOC, 10/14/82._______________________________ .__________ 6. In what ways is BFGoodrich using this information? The techniques are used to characterize complex polymer systems. 7. What is the patent situation? To our knowledge there is nothine in this report that is patentable. -Otf- *> 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Tooics. Boston. Mav 8-13. 1983. Publication in Macromolecules._____________ 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer._________________________________ 10 Oo Cl owetsivED JAN 131983 R.J. FAWCETT Signature of Author(s) U-,L C. 'cAc.^elcLosi (J sr-. ^__ JS t:. Signature of Author's Manager/ BG05^8 w. .x- *.w I Title of Paper Pyrolysis Studies of Lightly Chlorinated Poly(vinyl chloride) Author(s)______R. P. Lattimer, 'J. B. Pausch and H. L. C. Meuzelaar DateJanuary 5, 1983 Deadline Remarks 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pvrolvsis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for prese n/tation or publication? The annual.mass spectrometry conference is the best form for presentation of new results in this field. 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in Pyrolysis mass spectrometry (H. L. C. Meuzelaar) . 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The eeneral analytical techniques used have also been published. 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057;___________________f\ R. P. Lattimer, IOC, 10/14/82. \1 ^ l\ 6. In what ways is BFGoodrich using this information? The techniques are used to characterize complex polymer systemsX ......................................- ... - . - -T.i i'lr-' z 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 0729007 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Soectrometrv and Allied Topics. Boston. May 8-13. 1983.__ Publication in Macromolecules.; 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer.__________________________________ wrciveo JAN 131983 R.J. FAWCETT /1 ' jj VilUO^l___\i.LC ^Ar w^<2ACilLi Signature of Author(s) U I.T. Mlr Signature of Author1 s Manage^ BFG05149 Title of Paper Pyrolysis Studies of Lightly Chlorinated Poly(vinyl chloride) Author(s)R. P. Lattimer. J. B. Pausch and H. L. C. Meuzelaar DateJanuary 5, 1983 Deadline Remarks 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstruetural data (NMR) on CPVC. It is the first detailed analysis of a synthetic Polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? s The annual mass spectrometry conference is the best forum for presentation of new results in this field. 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in Pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published.______________________________________ _____________________________________________________ 5. To which Research Report does this paper relate? (Please Include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R. P. Lattimer, IOC, 10/14/82. 6. In what ways is BFGoodrich using this information? The techniques are used to characterize complex polymer systems. 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and__________ ^.3 Allied Topics. Boston. Mav 8-13. 1983.__ Publication in Macromolecules 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer.__________________________________ 006240 treeivD cs `JAN 131983 ^ R.J.FAWCETT ' [i.L C Signature of Author(s) <ff. Signature of Author's Managew' BFG05150 'Lj.cj.e 0 raper ryrinyaia omuida ui on mr mcicea rUi/i nnvi miut me;_________________________ Author (s)R. P. 'Lattimer. J. B. Pausch and H. L. C. Meuzelaar Dace February 2& 1983 Deadline Remarks tn be prpspntpd Mav 8-13. 1933____________ Abstract approved January, 1983. 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data v/ith microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference is the best forum for presentation of new results in this field._________________________________________________________ . 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The oeneral analytical techniques used have also ...been Published,:--------------------------------------------------------- ------ ------------------- _---------- 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R- P. I attimpr. TOC. 10/14/8?. R. P. lattimer and J. B. Pausch, Research_Report, 1/17/83. 6. In what ways is BFGoodrich using this information? The techniques are, used to characterize complex polymer systems. 7. What is the patent situation? To our know!edge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation bv Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. Mav 8-13. 1983. Publication in Macromplecules.------------------- __-----;---------- ___________________Rk'.t i w n 9. Was this paper requested or volunteered? By Whom? Volunteered " " by 11 Lattimer. " '" " i- . " . ___________________ ^ f/AR 3 _ i993 r-r * ^ " * * * * ** ^* *-- * j 0729003 BFG05151 Title of Paper Pyrolysis Studies of Chlorinated Polvfvinvl chloride) R.Author(s) P. Lattimer, J. B. Pausch and H. L. C. Meuzelaar_____________________________ Date February 28> 1983 Deadline Remarks to be presented May 8-13, 1983 Abstract approved January, 1983. 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. i 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference IS the best forum for presentation of new results in this field.______________________________ ; j 3. What in this paper will help BFGoodrich if presented or published? j: It will show continuing BF6 expertise in the development of new techniques for polymer i characterization. It will show BF6 collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? c | No technical advantage will be compromised. NMR structural characterization of these j materials has already been published. The general analytical techniques used have also : been published._____________________________ _____ !--------------------------------------------------------------- i 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057;________________________ R- p- I att.imer. TOC. 10/14/8?. R, P. I attimer and J. B. Pausch. Research Report, 1/17/83. j 6. In what ways is BFGoodrich using this information? : The techniques are, used to characterize complex polymer systems.. 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation bv Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. May 8-13. 1983. Publication in Macromolecules^--------------- -------------------------- ___________ Wtctivcu 9. Was this paper requested or volunteered? By Whom? fvlAR 3 _ l~-i3 Volunteered by Lattimer. i '_________________ ________________ r> i rravurcTT vs**?? Title of Paper Pyrolysis Studies of Chlorinated Polvtvinvl cnionoej 'Author(s)_______ R. P. Lattimer, J. B. Pausch and H. 1 . C. Meuzelaar____________ Dace February 28, 1983 Deadline Remarks to be presented Mav 8-13. 1983 Abstract approved January, 1983. 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It fs the first detailed analysis ofa synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference is the best forum for presentation of new results in this field.____________________ ;__________________________:______________ 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What Information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? Wo technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published.____________________________ __ _____________________ _______ 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25r4057; ______________ R. P. Lattimer. IQC. 10/14/82. R. P. Lattimer and J. B. Pausch, Research Report. 1/17/83. 6. In what ways is BFGoodrich using this information? The techniques are, used to characterize complex polymer systems 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 3. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. Mav 8-13. 1983. Publication in Macromolecules,----------------------------------- ________________________ > if o 9. Was this paper requested or volunteered? By Whom? Volunteered by Lattimer. ' _ MAR 3 _ 1333 Jia ! * **J -- * 20729011 Signature of Author(s) (4-lC U fi Signature of Author's Manager^ BFG05153 REQUEST FOR APPROVAL FOR PUBLICATION of Paper Pyrolysis Studies of Chlorinated Polvfvinvl chloride) Author(s) R. P. Lattimer. J. B. Pausch and H. i. C. Meuzelaar Date February 28 1983 Deadline Remarks to be prpspnted Mav 8-13, 1983 ________ Abstract approved January, 1983. L. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. ' 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference is the best forum for presentation of new results in this field._____________________________________________________________ 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymercharacterization. It will show BFG collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also Jbeen published.______________________________ ______________________________________________ 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R. P. I att.impr. TOC. 10/14/82. R. P. Lattimer and J. B. Pausch. Research Report, 1/17/83. 6. In what ways is BFGoodrich using this information? The techniques are, used to characterize complex polymer systems._________________________ 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable.^ --------------------------- :------- ---------------------:-----------------------------------------------:------------------------------------------- ----------------------O 3. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate^ where and when. Oral presentation by Lattimer at 31st Annual Conference on Mass Spectrometry and Allied jj^f -Topics. Boston. Mav 8-13. 1983. Publication in Hacromolecules, & j ' dwu v BFG05154  REQUEST FOR APPROVAL FOR PUBLICATION Title of Paper Pyrolysis Studies of Chlorinated Polyfvinyl chloride)_________________ Author(s)R. P. Lattimer. J. B. Rausch and H. L. C. Meuzelaar Date February 28 1983 Deadline Remarks to be presented May 8-13, 1983_________ Abstract approved January, 1983. 1. Why do you feel this paper is a worthwhile contribution to the technical literature? It correlates analytical pyrolysis data with microstructural data (NMR) on CPVC. It is the first detailed analysis of a synthetic polymer using pyrolysis-mass spectrometry in combination with computerized pattern recognition techniques. 2. Why is this time most suitable for presentation or publication? The annual mass spectrometry conference is the best forum for presentation of new results in this field.________________ _______________________ |_________________ ___ 3. What in this paper will help BFGoodrich if presented or published? It will show continuing BFG expertise in the development of new techniques for polymer characterization. It will show BFG collaboration with a recognized authority in pyrolysis mass spectrometry (H. L. C. Meuzelaar). 4. What information in this article decreases BFGoodrich technical advantages over competi tors and to what extent? No technical advantage will be compromised. NMR structural characterization of these materials has already been published. The general analytical techniques used have also been published. ------------------------------------------------------------------------------------------- -------------------- 5. To which Research Report does this paper relate? (Please include Access No.) R. P. Lattimer, Project 9101-81, 10/22/81, A.N. 25-4057; R. P. Lattimer. TOC. 10/14/82. R. P. Lattimer and J. B. Pausch. Research Report, 1/17/83. 6. In what ways is BFGoodrich using this information? The techniques are, used to characterize complex polymer systems. 20729013 7. What is the patent situation? To our knowledge there is nothing in this report that is patentable. 8. Do you plan on giving this paper orally? Do you propose to publish it? Please indicate where and when. Oral presentation bv Lattimer at 31st Annual Conference on Mass Spectrometry and Allied Topics. Boston. Mav 8-13. 1983. Publication in Macromolecule_s^----------------------------------------- ' ' --1 -- ' --IK -..I II ................ ' '` ' ' *-<, f % )-- >;; ?. ? C * Pyrolysis Studies of Chlorinated Poly(vinyl chloride) Robert P. Lattimer* and Jerry B. Pausch, The BFGoodrich Research and Development Center 9921 Brecksville Road, Brecksville, Ohio 44141 (U.S.A.) and ti Henk L. C. Meuzelaar, Biomaterials Profiling Center, University of Utah, 391 South Chipeta Way, Research Park, Salt Lake City, Utah 84108 (U.S.A.) ABSTRACT i ' ' ` The thermal decomposition of chlorinated PVC has been studied by pyroly sis-gas chromatography (Py-GC), Py-GC-mass spectrometry (Py-GC-MS), and Curie point pyrolysis-mass spectrometry (Py-MS). It was found that the yield of chlorobenzene pyrolyzate as determined by Py-GC correlates well with the weight-percent chlorine in the polymer or with microstructure parameters determined by 13C NMR. Py-MS data for several CPVC samples were studied by computerized pattern recognition techniques (factor analysis and canonical variate analysis). It was found that numerous Py-MS fragments (namely a number of - aromatic and chloroaromatic hydrocarbons) are strongly correlated with increases in the chlorine content of the polymer. Aliphatics and aromatics with more aliphatic character are negatively correlated with increasing chlor ine level in the polymer. Py-MS with computerized statistical analysis holds considerable potential for characterizing the composition and studying thermal decomposition mechanisms in synthetic polymers. BFG05156 INTRODUCTION The thermal decomposition behavior of chlorinated poly(vinyl chloride), CPVC, has been the subject of several studies., Berticat prepared CPVC with a. _ high chlorine level (73.2 wt.-% Cl) and found that CPVC was more thermally stable than PVC and that the principal volatile pyrplyzate was hydrogen chlor ide.1 In later work Berticat . .reported vthermogravimetric (TGA) results that .. showed a two-phase thermal decomposition for CPVC;(73-74% Cl).2 .The first : -i-V ; . i- ;i. . .'. - -?' - phase (~270 - 300C) was mostly dehydrochlorination, and the second phase at higher temperatures was characterized by the evolution of chlorine-containing hydrocarbons. The greater thermal stability observed for CPVC as compared to PVC was attributed in part to crosslinking reactions.2 .Tsuge et al. studied a `-i- series of CPVC's with chlorine contents up to 73.2% by pyrolysis-gas chromato graphy (Py-GC).3 The volatile pyrolyzates. detected at 460C were benzene, toluene, naphthalene, chlorobenzene, dichlorobenzenes (three isomers), tri chlorobenzenes (three isomers), and tetrachlorobenzenes (three isomers). Tsuge . et al. investigated the relative yields of these pyrolyzates as a function of the degree of chlorination. At "low" chlorine levels (up to ~62%), the rela tive benzene concentration decreased, while the chlorobenzene increased. The chlorobenzene concentration reached a maximum at ~64% chlorine,' and then decreased as di- and trichloro compounds became dominant. The Py-GC results were used to deduce microstructural information regarding the CPVC samples. In more recent work liebman et al. studied.the thermal decomposition and smoke evolution of PVC and CPVC. Electron spin resonance measurements demon strated that free radicals are present during dehydrochlorination.4 It was .suggested that the better stability of CPVC compared to PVC was due to long STOSZ tO Z BFG05157 sequences of -CHC1- ,units in the polymer.5 The better crosslinking (char forming) properties of CPVC as compared to PVC were also noted.5 Liebman et al. also studied the evolution of smoke from burning PVC and CPVC.6 A steady decrease in smoke evolution was observed as the chlorine level was increased. It was concluded that smoke was high in (low-chlorine) polymers because'these materials evolved larger quantities of benzene and chlorobenzene - both of which burn with a smoky flame. More highly chlorinated PVC's evolve di-, tri-, and tetrachlorobenzenes upon pyrolysis; these aromatics yield less smoke upon burning.6 Interpretation of results from earlier pyrolysis/combustion experiments3'6 was somewhat hampered by the lack of definitive microstructural data for the CPVC samples studied. 13C NMR has recently made great progress in defining the microstructural units in CPVC and related polymers.7-9 We have examined by Py-GC and Py-MS a series of chlorinated PVC's that have also been characterized by 13C NMR.9 Our primary objective was to determine if correlations could be made between the NMR and pyrolysis data. A secondary objective was to deter mine if Curie point Py-MS in combination with pattern recognition techniques could differentiate CPVC samples with varying chlorine contents. This report describes our preliminary results and conclusions. BFG05158 20729016 EXPERIMENTAL CPVC samples. The chlorinated PVC's (< 61 wt.-% Cl) were prepared by solution photochlorination of Geon 103EP PVC (The BFGoodrich Company) in tetrachloroethane at 80C. The samples and numbering system are the same as described in the NMR study..9 Py-GC and Py-GC-MS. In our study five of the seven NMR samples9 were analyzed by pyrolysis-gas chromatography (Py-GC). NMR characterization data along with chlorobenzene pyrolyzate yields are given in Table I. 40 pg of polymer was deposited on the ribbon pyrolysis probe from tetrahydrofuran solution. After evaporation of the solvent, the samples were pyrolyzed at 600C for 20 s (temperature rise time ~8 ms) in helium atmosphere using a Chemical Data Systems Model 100 Pyroprobe. The pyrolyzates were separated with a 40 m Carbowax 20M fused silica capillary column. A Varian 3700/CDS 111 GC-microprocessor system (The BFGoodrich Company) was used with flame ionization detector. The capillary injector split ratio was 50/1. Calibration was achieved using injections of chlorobenzene in n-pentane solution (external standard method). Qualitative identification of pyrolyzates was achieved by Py-GC-mass spectrometry, using a 40 m Carbowax 20M glass capillary column for separation. A Varian 3700/Finnigan MAT 311A/Finnigan Incos 2400 GC-MS-DS (The BFGoodrich Company) was used. Figure la shows the mass pyrogram for sample 1 (PVC), and Figure lb shows sample 7 (CPVC). A trace of chlorobenzene appeared in sample 1, while several chloroaromatics appeared in sample 7 (chlorobenzene is the largest). BFG05159 20729017 Py-MS. Six of the NMR samples9 were also analyzed by Curie point Py-MS using an Extranuclear 5000-1 system (The University of Utah). About 4 pg of polymer was applied by microsyringe to the pyrolysis wire from tetrahydrofuran solution and air-dried under gentle rotation of the wire. Each sample was analyzed in triplicate. Py-MS conditions were as follows: Curie point temperature 6l0C, temperature rise time 5 s, total heating time 10 s, reaction tube directly in front of the ion source (no expansion chamber used), electron energy setting 11 eV, mass range scanned m/z 25-240, scan rate 2000 amu/s, total scan time ~25 s. For each sample a single integrated spectrum was recorded. Spectra for sample 1 (PVC, Figure 2a) and sample 7 (CPVC, Figure 2b) are included. Py-MS data processing. Computerized processing of the Py-MS data involved the use of spectrum calibration and normalization procedures described elsewhere,10 as well as multivariate statistical analysis routines available in the Statis tical Package for the Social Studies (SPSS).11 Multivariate analysis consisted of principal component analysis of the complete Py-MS data matrix followed by canonical variate analysis of the correlations between the nine most signifi cant (eigenvalues > 1.0) principal components (factors) and the chlorine content of the CPVC samples as determined by 13C NMR.9 Finally, the canonical variate "spectrum" was calculated according to a procedure described by Windig et al.12 BFG05160 20729018 RESULTS AND DISCUSSION Chlorobenzene pyrolyzate yield. It was of interest to determine whether the yield of chlorobenzene as a pyrolyzate could be used as an indicator of the level of chlorine in CPVC. Only a trace of chlorobenzene is observed as a pyrolyzate in pyrolysis gas chromatography of PVC itself (Figure 1 and Table I). As the amount of chlorine in CPVC increases, the yield of chlorobenzene pyrolyzate, as detected by Py-GC, increases in a regular manner (Table I). The correlation of chlorobenzene pyrolyzate yield with Schoniger %-Cl is not at all good, since the Schoniger method is only accurate to ~0.5%. Chlorobenzene correlation with %-Cl by 13C NMR is nearly linear (Figure 3, correlation coefficient .988), but a better linear correlation is obtained by plotting chlorobenzene yield versus mole-% (111) + (020) structures from NMR (Figure 4, correlation coefficient .999). This linear correlation is consistent with established PVC decomposition pathways,13-15 since either (111) or (020) structures can lead to chlorobenzene formation: 20723019 -V- -V-ch2- chci -chci - CHCI -ch2 - CHCI (011101) A`1I--3HCI -AA-CH * CH- CCI = CH -CH = CH -AA- CYCL1ZATI0N \/ H-/ '"Sj-h Vl " IV HH 01 H L h CHL0R08ENZENE -AA-CH2-CHCI-CH2-CCI2 -ch2 -chci-AA- (010201) --3HCI AA--Af^-- CH = CH -- CH = CCI --CH = CH-- CYCLIZATION W' // H--C \ C--H W jTl l-K CLEAVAGE HH CHLOROBENZENE 20729020 BFG05162 *$** We would expect this linear relationship to hold up to the point at which dichlorobenzene formation starts to become important compared to chlorobenzene. This transition is apparently ~61-62%-Cl.3 Py-MS correlations. The largest pyrolyzate peaks in the Curie point pyrolysis s> + ^ mass spectra (Figures 2 a,b) were m/z 36-38 [HC1] * and m/z 78 [CeHg] *. Ions m from residual tetrahydrofuran solvent (m/z 72-71-42) were also prominent in most samples; these peaks were ignored in the' pattern recognition analyses. Some samples also showed appreciable [S02]+* (m/z 64) from an unknown source. The regular increase of chlorobenzene yield with increasing %-Cl is quite apparent in the Curie point Py-MS data. A bivariate plot of m/z 112 [CgH53SCl]+* versus m/z 114 [C6H537C1]+* is given in Figure 5 (correlation coefficient .995). Note that appreciable intensities of m/z 112 and 114 were observed even for sample 1 (PVC), which yields only a trace of chlorobenzene pyrolyzate. These "baseline" intensities are due mainly to small amounts of hydrocarbon pyrolyzates (C8H16 and CgH18). A plot of the relative intensity of (m/z 112 + m/z 114) versus mole-% (111) + (020) is shown in Figure 6. The graph shows some curvature (correla tion coefficient .943). This nonlinearity is apparently due to the fact that relative, rather than absolute, intensities of (m/z 112 + 114) have been plotted. Since the overall yield of volatile pyrolyzates in CPVC decreases as the chlorine level increases,6 we should not expect Figure 6 to give a good linear correlation.3 20729021 As known from the literature3 as well as evidenced by the chromatogram in Figure lb, several chloroaromatic compounds in addition to chlorobenzene can be found as CPVC pyrolyzates. Since the intensities of such chloroaromatic species in the pyrolysis mass spectra of CPVC samples may be expected to correlate more or less strongly with chlorine content (as shown in Figures 5 and 6 for chlorobenzene), multivariate statistical analysis methods10-12 can be // used to identify these correlated peak series. Figure 7 shows the correlation between the canonical variate function calculated from the Py-MS data and the chlorine content values derived from NMR data.9 As described in the Experimen tal section, this canonical variate function was calculated in the space described by the nine most significant principal coordinates (factors) of the Py-MS data using the canonical variate routine in the SPSS program.11-12 Only peaks in the range m/z 78-177 were included in this calculation because of limitations in the size of the data matrix which can be handled by the program. The high correlation coefficient (0.995) and the sizable fraction (34.0%) of total variance in the Py-MS data set explained by the canonical variate func tion demonstrate that differences in chlorine content are among the most pronounced tendencies in the spectral patterns. Because of the very limited size of the present set of samples no attempt was made to test the reliability of the canonical variate function for indicat ing the chlorine content of unknown samples. However, apart from its potential usefulness as a quantitative tool, the canonical variate function also enables the visualization of ion signals most strongly correlating with changes in chlorine content. Figure 8, shows a canonical variate spectrum calculated according to Windig et al.12 The spectrum is dominated by chlorobenzene signals (m/z 112-114), but several other chloroaromatic signals are apparent at BFG05164 20729022 iU m/z 126-128 (chlorotoluene), m/z 138-140 (chlorostyrene), m/z 150-152 (chloroindene), and m/z 162-164 (chloronaphthalene); most of these were also observed by Py-GC-MS (Figure lb). In addition, a pronounced series of nonchlorinated aromatic pyrolyzates is also present - m/z 78 (benzene), m/z 92 (toluene), m/z 104 (styrene), m/z 106 (C2 alkylbenzenes), m/z 116 (indene), m/z 128 (naphthalene), m/z 142 (methylnaphthalenes), m/z 154 (biphenyl/acenaphthene), m/z 156 (C2 alkylnaphthalenes), m/z 166 (fluorene), and m/z 168 (methylbiphenyls/ methylacenaphthenes). A final peak that shows a positive correlation is m/z 87, which we believe is an artifact due to degraded (oligomerized) tetrahydrofuran solvent. Several hydrocarbon peaks are also observed on the negative side of the canonical variate spectrum (Figure 8). These compounds are either unsaturated aliphatics or else aromatics with appreciable aliphatic character. It is probable that lower molecular weight aliphatic hydrocarbons would also appear on the negative side of the spectrum, but the canonical variate analysis did not include mass numbers less than m/z 78. Overall our results suggest that the CPVC pyrolyzates can be divided into three categories with respect to their formation tendencies: 1. Chloroaromatic compounds increase in both relative and absolute abundance as the chlorine level in the polymer increases. This was documented in earlier work3'6 and has been confirmed in our studies. 2. Pure conjugated aromatic hydrocarbons and aromatics with small (Cx - C2) aliphatic side groups increase in relative abundance as the chlorine level BFG05^65 II 20729023 increases. Other work shows that in absolute abundance these compounds actually decrease with increasing chlorine level.3'16 3. Aliphatics and aromatics with more aliphatic character decrease in both, relative .and absolute abundance with increasing chlorine .'level .in ...the polymer. Our results show that as the chlorine level in the polymer increases, the volatile hydrocarbon pyrolyzates exhibit an overall decrease ..in hydrogen-tocarbon ratio. There are three possible explanations for this decrease in hydrogen in the volatile pyrolyzates: (1) there is less hydrogen available in the polymer; (2) more hydrogen is used to form HC1; and .(3) more hydrogen remains in the polymer residue as char. Explanation (1) is certainly operative for CPVC, since for every chlorine atom added to the polymer one hydrogen atom is lost. Explanation (2) is also operative, since presumably more HC1 is formed from the polymer during pyrolysis as the chlorine level increases. The HC1 pyrolyzate yield as a function of polymer chlorine level has apparently not been the subject of careful study, however. Explanation (3) is probably not valid, since chars are generally very deficient in hydrogen. While PVC itself produces little char during thermal decomposition, addition of chlorine to the polymer results in a steady increase in char residue.6'17 This implies that crosslinking mechanisms not significant in PVC become important as more chlorine is added to the polymer. These mechanisms may involve intermolecular HC1 elimination, Diels-Alder reactions, and various free radical processes. N O' Nl*0 to O' N These preliminary results suggest that Py-MS with computerized statistical analysis holds considerable potential for characterizing the composition and studying thermal decomposition mechanisms in chlorine-containing polymers. Further studies are in progress that should help to elucidate the thermal decomposition behavior of CPVC in more detail. s* . Acknowledgment. Appreciation is expressed to The BFGoodrich Company for support of this work. W. Vindig and A. M. Harper (University of Utah) assisted with the computerized statistical analysis, and W. H. McClennen (University of Utah) performed the Curie point Py-MS experiments. I. Sockis and R. E. Harris (The BFGoodrich Company) assisted with the Py-GC and Py-GC-MS experiments. Helpful discussions with E. D. Dickens, Jr., G. S. Huvard, R. A. Komoroski, W. J. Kroenke, M. H. Lehr, and R. G. Parker (The BFGoodrich Company) are appreciated. S. A. Liebman (Chemical Data Systems, Inc.) is also acknowledged for encouraging us to work on this project. BFG051&'7 20729025 REFERENCES AND NOTES 1. Berticat, P. J. Chim. Phys. 1967, 64, 887. 2. Berticat, P. Rev. Gen. Caout. Plast. 1971, 48, 1361. // 3. Tsuge, S.; Okumoto, T.; Takeuchi, T. Macromolecules, 1969, 2, 277. r* 4. Liebmaa, S.A.; Reuwer, J.F., Jr.; Gollatz, K.A.; Nauman, C.D. J. Polym. Sci. A-l, 1971, 9, 1823. 5. Liebman, S.A.; Ahlstrom, D.H.,; Quinn, E.J.; Geigley, A.G.; Meluskey, J.T. J. Polym Sci. A-l, 1971, 9, 1921. 6. Quinn, E.J.; Ahlstrom, D.H.; Liebman, S.A. Polym. Preprints, 1973, 14(2), 1022. 7. Lukas, R,; Svetly, J.; Kolinsky, M. J. Polym. Sci., Polym. Chem. Ed., 1981, 19, 295. 8. Keller, F.; Hosselbarth B. Faserforsch. Textiltech., 1978, 29, 152. 9. Komoroski,. R.A.; Parker, R.G.; Lehr, M.H. Macromolecules, 1982, 15, 844. 10. Harper, A.M.; Meuzelaar, H.L.C.; Given, P.H.; Pope, D.L.; Metcalf, G.S. "Analytical Pyrolysis"; Voorhees, K.J., Ed.; Butterworth: Voburn, MA, 1983; in press. b?G05^68 20729026 11. Nie, N.H.; Hull, C.H.; Jenkins, J.G.; Steinbrenner, K.; Bent, W.H. "Stat istical Package for the Social Sciences"; 2nd Ed.; McGraw-Hill: New York, 1975. 12. Windig, W.; Meuzelaar, H.L.C.; Haws, B.A.; Campbell, W.F.; Asay, K.H. J. Plant Diseases Protection;^in press. 13. O'Mara, M.M. Pure Appl..Chem.t 1977, 49, 649. 14. Starnes, W.H., Jr.; Edelson, D. Macromolecules, 1979, 12, 797. 15. Lattimer, R.P., Kroenke, W.J. J. Appl. Polym. Sci., 1982, 27, 1355. 16. Lattimer, R.P., unpublished results. 17. Dickens, E.D., Jr., The BFGoodrich Co., private communication. b?G05^69 20729027 TABLE I. Chlorobenzene Pyrolyzate Yield and Other Characterization Data for Lightly Chlorinated PVC's. . Sample No. 1 3 4 5. .6 : 7 '- . % Cla % Clb (NMR) mol %b (111) mol %b (020) wt.-%c.- Chlorobenzene ; - 56.2 56.7 56.8 - 57.3 >>- 57.8 57.7 ^ 58.3 v . 58.7 "V? 58.0 ; V 59.5 57.9 - 7 ;7?60.1 59.4 ` " i- ' _ i . 760.6 V i 0.0 0.0? - l.l v.ro.o ~0.d . ' ; "o.osi+o;oi2 - 1.7 . 0.0 0.1510.04 7 7 :3.3 0-5 . ' ,.. nde . 7- 4.o ; :'0.5 nd '' ;2": .6 ' ^0.6 . ,0.4810.07 .%*?> V 7 >-7.1 : 7 1.4 '~77i o.69o:67 377:7. , 7- ' V>77i.y'"j^rrZTV'':''r*- .j a Schoniger oxygen flask method (data repeated from reference -9). b 13C NMR (from reference 9). . - i .. . ' ' \';-r ' c Pyrolyzate yield (based on weight of polymer). Standard deviations are given from multiple runs. d Trace of chlorobenzene observed by Py-GC-MS. 1. e Not determined. 20729028 BFGOSHO Pyrolysis Studies of Lightly Chlorinated Poly (vinyl chloride) by R. P. Lattimer and J. B. Pausch The BFGoodrich Research and Development Center Brecksville, Ohio 44141 and // H. L.C. Meuzelaar Biomaterials Profiling Center, University of Utah Salt Lake City, Utah 84108 ABSTRACT The thermal decomposition of lightly chlorinated PVC has been studied by pyrolysis-gas chromatography (Py-GC), Py-GC-mass spectrometry, and Curie point Py-MS. It was found that the yield of chlorobenzene pyrolyzate as determined by Py-GC correlates well with the weight-percent chlorine in the polymer or with microstructure parameters determined by 13C NMR. Py-MS data for several CPVC samples were studied by computerized pattern recognition techniques (factor analysis and canonical variate analysis). It was found that numerous Py-MS fragments (namely a number of aromatic and chloroaromatic hydrocarbons) were strongly correlated with changes in chlorine content in the polymer. Py-MS with computerized statistical analysis holds considerable potential for characterizing the composition and studying thermal decomposition mechanisms in synthetic polymers. BFG05VH 20729029 Dr. D. E. Ley_________________Director, Corporate Technical Support FROM S'jajECT FIELD POINT OR DEPT. & BLOG. NO. Research & Development Center- Brecksville PUBLICATION APPROVAL DATE THIS LETTER 3-17-83 - The BFGoodrich Publications Committee has approved for publication the paper entitled "Pyrolysis Studies of Chlorinated Poly(vinyl chloride)" by Messrs. R. P. Lattimer, J. B. Pausch and H. L. C. Meuzelaar. Our records indicate that the paper will be presented orally by Dr. Lattimer at the 31st Annual Conference on Mass Spec trometry and Allied Topics in Boston, Massachusetts, May 8-13, 1983. And the paper will be published in Macromolecules. P7^ couML&t/tf \ R. J2 Fawcett d ends. i. cc: Research .Files, Brecksville - w/Manuscript and approval sheets Public Relations, Akron - w/Manuscript N. w. Shust - w/Manuscript 20729030 s,?'i.4SSS- 1/60 LITHO IN U.S.A. BFG05V72 The BFGoodrich Company Research and Development Center 9921 Brecksville Road Brecksville. Ohio 44141 216-447-5201 Robert J. Fawcett Vice President Research and Engineering March 17, 1983 Dr. Robert P. Lattimer R&D_ Associate Corporate Technical Support Brecksville, Ohio Dear Bob, I am pleased that the paper entitled "Pyrolysis Studies of Chlorinated Poly(vinyl chloride)" that is co-authored with Messrs. J. B. Pausch and H. L. C. Meuzelaar has been approved by the BFGoodrich Publications Committee. Our records indicate that you will present the paper orally at the 31st Annual Conference on Mass Spectrometry and Allied Topics in Boston, Massachusetts, May 8-13, 1983. And the paper will be published in Macromolecules. I wish to congratulate you on a fine piece of work correlating analytical pyrolysis data with the microstructure of CPVC. To my knowledge, this is the first detailed analysis of a synthetic poly mer using the pyrolysis-mass spec in combination with computerized pattern recognition techniques. The results of this work should aid in determining methods for producing more fire resistant, low smoke CPVC resins. This collaboration of industrial and academic scientists should be of interest to many. Sincerely RJF/d 20729031 BFG05173 ! i i' i 1 The BFGoodrich Company Research and Development Center 9921 Brecksville Road Brecksville. Ohto 44)4) 216-447-5201 Robert J. Fawcett Vice President Research ond Engineering / / Dr. Jerry B. Pausch R&D Manager Corporate Technical Support Brecksville, Ohio March 17, 1983 Dear Jerry, I am pleased that the paper entitled "Pyrolysis Studies of Chlorinated Poly(vinyl chloride)" that is co-authored with Messrs. R. P. Lattimer and H. L. C. Meuzelaar has been ap proved by the BFGoodrich Publications Committee. Our records indicate that the paper will be presented orally by Dr. Lattimer at the 31st Annual Conference on Mass Spectrometry and Allied Topics in Boston, Massachusetts, May 8-13, 1983. And the paper will be published in Macromolecules. I wish to congratulate you on a fine piece of work correlating analytical pyrolysis data with the microstructure of CPVC. To my knowledge, this is the first detailed analysis of a synthetic polymer using the pyrolysis-mass spec in combination with computerized pattern recognition techniques. The results of this work should aid in determining methods for producing more fire resistant, low smoke CPVC resins. This collaboration of industrial and academic scientists should be of interest to many. Sincerely, ~^3u&1r RJF/d 20729032 BFG05174 V. BFG05175 Figure 2. CURIE POINT PYROLYSIS MASS SPECTRA. a. SAMPLE I (PVC) b. SAMPLE 7 (CPVC) 20729034 4'^Jr.! bfgo5V76 ?Sr**^re^C" CHLOROBENZENE (Py-GC) $t06Z0Z Figure 3. CHLOROBENZENE' PYR0UATE YIELD (Py-GC) versus W*V BFG05177 Figure 4. CHLOROBENZENE PYROLYZATE YIELD (Py-GC) versus MOLE PERCENT (lll)+(020). -MOLE % (III) +1020) 20729036 BFG05178 Figure 5. BIVARIATE PLOT (Py-MS) of m/z 112 and m/z 114. BFG05179 20729037 t 20729038 Figure 6. MOLE-PERCENT (lll) + (020) versus (m/2 112+m/z 114). B FG05180 Figure 7. CANONICAL VARIATE SCORE versus PERCENT CHLORINE. (CHLORINE l3C NMR) 20729039 BFG05181 20729040 'I Figure 8. CANONICAL VARIATE SPECTRUM. 0.7H- T N- o o (SlINn *9dV) AJLISN31NI BFG05182 /d ' 6FG TECHNICAL DOCUMENT)BFG TECHNICAL DOCUMENT INTERNAL BFG USE ONLY 4u4mO* OATt KgrORT i.o tote mcrcigNcr J. 11. I/IWDHAY <huitto it ALTC/PSC 1/23/84 ciiwsr------------- Mditct mo. 002-147-101 SOM* ritt SCO mo. R. O. HARDSSTY ALTC/416 4064-0100 MALBIC AKHYDRXDR MONITORING POR NSW BROW GEON COMPOUND Ijf.V"-*- i ' rititococTiCM-saiWMBr A nmm brown Gon* oompound is being developed for use as a Window frama profile. Tbe stabiliser used, tin bis(butyl maleate) has [tjkm potential to generate naleie anhydride during compounding. H'ipribM"tiaH. terlag this development, mill roam operators as mall as eye and respiratory irri- 1| determine if maleie anhydride (H.A.) could be the tense of those discomforts air samples mere monitored a oomopumdlng (1/11-12/84) and an extrasion run (1/18/M). imp tet hteo saaples submitted for M.A. analysis mere over tbe W of l.O ntel with the highest levels CoonA at the *< mill rotf# t1>Hat/fc3) 1/11/M and: at the Untwiry drop pan (40.7 .iteiff*)' 1/U/M. All- further production runs of tbe brown Geon >> i as Binding future mork on , interiellf ventilated.d> * sns prahlahs mere ml and Tf . aii llguid (2| by the procedure an pLes analysed and the results obtained. ,,rmis detected, the minimum detectable hraefceta. These KDL values vary due to S end air voltanes. Samples 42 and 4182 tbs TLV of 1.8 mg/ml. The high minimum il 41 (<3.28) maa obtained due to M4 V' BFG05183 !H DOCUMENT NUMBER: Pulled for Attorney-Client Privilege / Work Product THE FUNCTION OF MOLYBDENUM ADDITIVES IN THE THERMAL DECOMPOSITION OF POLY(VINYL CHLORIDE) BY WILLIAM J. KROENKE TWe BFGOODRICH COMPANY RESEARCH AND DEVELOPMENT CENTER BRECKSVILLE, OH Abstract Molybdenum compounds are a versatile class of smoke and fire retard er additives for poly(vinyl chloride), PVC. They act to change the thermal decomposition pathways of PVC. Although PVC is naturally fire retardant, when it is forced to burn, it generates smoke and leaves behind only a small residue. Molybdenum additives, in contrast, signifi cantly reduce smoke formation and promote the formation of char which decreases the amount of PVC available as fuel. [This talk presents the results of recent research studies aimed at determining how molybdenum acts to change the thermal decomposition pattern of PVC. Smoke and char formation during combustion will be related to the products formed during inert atmosphere pyrolysis. Gas chromatography-mass spectrometry-pyrolysis studies were used to charac terize the volatile pyrolyzates. Pyrolysis studies were made using heterotactic, perdeuterated, and 13C2 PVC compounds. Synthesis and use of 2,4,6-trichloroheptane as a model compound for PVC provided unexpected insight into the molybdenum promoted PVC decompo sition mechanism. Precursors to char were observed for the first time. The results were also inconsistent with the Lewis acid and reductive coupling mechanisms respectively proposed by researchers at Bell Labora tories and BFGoodrichJ WJK/blr BFG05186 xmzsLoz TO ______Dr. Geoffrey A. Lindsay FROM SUBJECT PUBLICATION APPROVAL FIELD POINT OR DS>T. & BLDG. NO. Manager, Corporate Research - Brecksville DATE YOUR LETTER ________________ FIELD POINT OR DEPT. & BLDG. NO. DATE THIS LETTER Research & Development Center- ----------------------------------------' .......... ]_" '" ' 1 ~-----:--'}-'t ' The BFGoodrich Publications Committee has approved for publication the paper entitled "The Function of Moly bdenum Additives in the Thermal Decomposition of Poly(Vinyl Chloride). Our records indicate that the paper was presented orally at the Chemistry Colloquia Series at Case Western Reserve University on February 10, 1983. Ccuj-c&/ FaWcett d ends. RETENTION CODE 801-001 PROJECT REPORT PERMANENT RECORD COPY cc:Research Files, Brecksville - w/Manuscript and approval sheets Public Relations, Akron - w/Manuscript E. G. Fiorito - w/Manuscript C. E. Wilkes - w/Copy of approval sheets THE B.F. GOODRICH CO FEB 21 1983 RESEARCH LIBRARY 8FG-4956-E 11/80LITHO IN U.S.A. BFG05187 THE FUNCTION OF MOLYBDENUM ADDITIVES IN THE THERMAL DECOMPOSITION OF POLY(VINYL CHLORIDE) BY WILLIAM J. KROENKE THE BFGOODRICH COMPANY RESEARCH AND DEVELOPMENT CENTER JBRECKSVILLE, OH ABSTRACT Molybdenum compounds are a versatile class of smoke and fire retard er additives for poly(vinyl chloride), PVC. They act to change the thermal decomposition pathways of PVC. Although PVC is naturally fire retardant, when it is forced to burn, it generates smoke and leaves behind only a small residue. Molybdenum additives, in contrast, signifi cantly reduce smoke formation and promote the formation of char which decreases the amount of PVC available as fuel. This talk presents the results of recent research studies aimed at determining how molybdenum acts to change the thermal decomposition pattern of PVC. Smoke and char formation during combustion will be related to the products formed during inert atmosphere pyrolysis. Gas chromatography/mass spectrometry/pyrolysis studies were used to charac terize the volatile pyrolyzates. Pyrolysis studies were made using heterotactic, syndiotactic, perdeuterated, and 13C labeled PVC compounds. Synthesis and use of 2,4,6-trichloroheptane as a model compound for PVC provided unexpected insight into the molybdenum promoted PVC decompo sition mechanism. Precursors to char were observed for the first time. The results were also inconsistent with the Lewis acid and reductive coupling mechanisms respectively proposed by Starnes and Edelson3 and Lattimer and Kroenke.10 INTRODUCTION Fire safety is a universal concern. As outlined in a recent feature article in Chemical and Engineering News,1 there is no universal agree ment on the relative fire hazards of natural materials compared to synthetic materials. During combustion, both of these materials form smoke and toxic gases. The essential question, which remains unanswered, is whether the increased use of synthetic materials is resulting in an increased fire hazard? The issues are very complex. While it is unanimously agreed that carbon monoxide is the number one toxic gas hazard in the combustion of most materials, natural or synthetic, the role of other toxic gases specific to certain materials is much less clear, and is surrounded by controversy. BFG05188 20732003 The situation with regard to smoke is better understood and much less controversial. Smoke obscures vision and impairs a persons ability to flee a fire situation. This obscuration of vision tends to promote hysteria, which in turn results in misorientatipn and panic. While smoke is a product of most combustion processes and large scale fires, under similar laboratory test conditions some synthetic materials generate more smoke than wood. However, the ranking of materials in terms of smoke generation is very dependent on the specific test conditions. For example, under smoldering conditions, wood burns with the generation of copius quantities of smoke. In contrast, it generates little smoke under flaming conditions where it burns with a rapid flame spread. PVC behaves just the opposite of wood. When it is forced to burn, it generates little smoke under smoldering conditions, but significant smoke under flaming conditions. However, even in the flaming mode, it has a rate of flame spread significantly less than that of wood. The objective of the research reported here was to further reduce the already low flammability and combustibility properties of PVC. Special emphasis was placed on reducing the smoke generated from PVC when it is forced to burn, and in reducing its toxic gas hazard by reducing the amount of PVC which behaves as, or generates, a fuel for combustion. The approach selected was to find chemistries which would alter the normal thermal degradation pathways of PVC and promote the formation of a thermally stable char. The first part of this talk concerns the thermal decomposition pathways of PVC. New mechanisms for the formation of volatile aromatic pyrolyzates will be presented. These are consistent with the results from deuterium and *^C labeling studies. In the second part of this talk the development of appropriate smoke and fire retarder systems for PVC will be discussed. Particular emphasis will be placed on identifying the most effective additive systems, and the problems involved in developing smoke and fire retarder systems for commercially useful PVC compounds. ' The third part of this talk deals with the role of the metal-based smoke and fire retarders in reducing the formation of smoke and promoting the formation of char. Molybdenum based smoke and fire retarder addi tives will be featured. More specifically, the discussion will center around the mechanism(s) by which molybdenum trioxide (Mo03) acts as a smoke and fire retarder for PVC. The results of pyrolysis, gas chroma tography, mass spectroscopy (PY-GC-MS) experiments performed on a wide variety of different PVC compounds will be discussed. EXPERIMENTAL The model rigid PVC compound is 100 parts Geon 103EP resin, 2 parts microthene 510 polyethylene, and 2 parts dibutyltin bis(isooctylthioglycolate). The Geon resin is a homopolymer with an inherent viscosity of 0.98-1.04 and an ASTM classification of GP-5-1443. The microthene 510 is a processing aid while the tin thioglycolate is a stabilizer. V BFG05189 20732004 3 The PVC samples containing the candidate smoke retarders were prepared by milling at about 160C on a rolling rubber mill. Sheets of appropriate thickness were molded under pressure at about 165C. Smoke evolution primarily was determined by means of the NBS Smoke Chamber test. The PVC samples measured 7.3x7.3x0.06 cm and were burned in the flaming mode in accordance with ASTM-E-662-79 "Test for Specific Optical Density of Smoke Generated by Solid Materials". In this article, the NBS smoke number is expressed as D /g which is the smoke generated per gram of sample. In certain cases, smoke evolution was determined by both the NBS Smoke Chamber test and the Goodrich Smoke-Char test. The char-forming characteristics of certain of the polymer samples also were determined using the Goodrich Smoke-Char test. This smoke-char test is a small scale laboratory test which is useful for quickly evaluating the smoke-forming and char-forming charac teristics of polymer samples. Small (0.3-0.4 gm) polymer samples measur ing about 1.3x0.95x0.19 cm are placed on a screen and forced to burn by being totally immersed in the flame from a propane torch (276 Pa) for 1 min. The smoke from the burning samples rises through a vertical chimney and passes through the beam of a photometer. The photometer is coupled with an integrator which provides a measure of the total smoke evolved. The smoke number, SPVC, is expressed as integrated area per gram of PVC in the polymer sample. The residue of "char" remaining after the smoke-char test is weighed and used to calculate the "percent of backbone char" (% BC). Essential ly, % BC represents the amount of the PVC hydrocarbon backbone which has been retained as a thermally stable char. It is calculated as shown: % BC char wt - nonburnable residue wt sample wt - non-PVC weight - HC1 wt x 100 A CDS model 100 Pyroprobe was used. Platinum ribbon or coil probes were used when appropriate. Typical pyrolyses were carried out for 20-30 sec at 550-600C; the Pyroprobe temperature had been calibrated by CDS (optical pyrometer). The pyrolyses were conducted in dry helium environment at a flow rate of ~30 cm2/min. Polymer samples were deposit ed on the ribbon via syringe from tetrahydrofuran solution. The solvent was allowed to evaporate in air before pyrolysis was carried out. The pyrolyzates were analyzed by gas chromatography/mass spectro scopy. The GC/MS system consisted of a Varian 3700 Digital gas chromato graph, a Varian MAT (now Finnigan MAT) 311A mass spectrometer, and a Finnigan ENC0S 2400 data system. Typical system temperatures were as follows: PY-GC interface (injector), 150C; GC-MS interface, 290C; ion source, 220C. Pyrolyzates were separated on a 4 m x 2 mm ID glass 3% Dexsil 300 on 100/120-mesh Supelcoport column, programmed from 50-250C at 15 deg/min after a 2-min hold on injection. Most of the PY-GC-MS runs were made in El mode (~38 eV) using a combination chemical ionization/electron impact ion source. For the BFG05190 20732005 i 5 cooled in dry ice. The 13C2*VCM was added by injection the gas under the surface of the cold liquid. Using a Pyrex capillary, nitrogen was bubbled slowly through the mixture. After one minute, the tube was sealed with a torch. The nitrogen purge was continued until the flame collapsed the capillary. The sealed tube was tumbled for 16 h at 50C. The tan, waxy product was dissolved in tetrahydrofuran and precipitated with methanol. Yield of polymer was 59 mg (15%). control sample was prepared under the same conditions using 12C2_VCM; the yield of polymer (a white powder) was 62%. VOLATILE AROMATIC PYROLYZATE FORMATION FROM POLY(VINYL CHLORIDE) - -* Introduction ' ^ A fundamental goal in this study was to determine the basic mechan isms by which volatile aromatic pyrolyzates form from PVC in the absence of additives. As will be shown in a latter section of this talk, the combustion of aromatic pyrolyzates is the principle source of smoke from burning PVC. Furthermore, the effect of smoke retarders we have devel oped for PVC are known to interfere with the normal mechanisms by which volatile pyrolyzates form. Thus we hoped a good understanding of pyro lysis mechanisms in virgin PVC would provide insight into the mechanisms by which smoke-reducing and char-forming additives work in PVC compounds. Our earlier PY-GC-MS studies in this area used GC-field ionization mass spectroscopy (FI-MS) to analyze the volatile pyrolyzates.2 The use of FI-MS greatly simplified the spectral interpretation in deuterium tracer experiments, since essentially the only ions observed were mole cular ions. Unfortunately, however, GC-FI-MS is not a very sensitive technique compared to GC-electron impact mass spectroscopy (EI-MS). In the first studies, we used relatively large quantities of PVC (~2 mg) in the pyroprobe in order to obtain sufficient amounts of pyrolyzate for adequate FI-MS detection. However, we found that secondary pyrolysis reactions were enhanced by the relatively large quantities of PVC and gave more H/D mixing than would be expected from primary pyrolyzate formation mechanisms. The use of EI/MS at normal electron energies (~70 eV) gives optimal sensitivity for detection. However, when a mixture of isotopic species is present, fragmentation reactions lead to very complex mass spectra that are difficult to interpret. For example, in our first pyrolysis studies of mixtures of PVC and pedeutero-PVC (DPVC), all isotopic species of toluene from C7H8Do to C7H0D8 were observed. An alternative method of analysis, which provides a compromise between FI-MS and 70-eV EI-MS, is lowered voltage EI-MS. At lowered voltage the sensitivity for detection is intermediate between EI-MS and FI-MS. The mass spectra are simpler than 70-eV El spectra but more complex than the simple mplecular ion FI spectra. In this work we have repeated our PY-GC-MS experiments with intimate mixtures of PVC and DPVC using smaller quantities of polymer in the pyroprobe (~40 pg compared to ~2 mg). The pyrolyzates were analyzed by 20732006 BFG05191 7 steps.4 The cyclohexadiene intermediate (II) seems reasonable based on consideration of activation energies.5 The conversion of II to III is likely a two step process, since a concerted reaction is not allowed by orbital symmetry.3 Also, the presence of macroradical intermediates during PVC decomposition has been verified by ESR experiments.6 Finally, the calculated activation energies for conversion of II to III via a two-step process seem consistent with the observed elimination of benzene at relatively low temperature/s/. Toluene Formation Two distinctly intermolecular pathways for toluene formation are depicted in Figure 3. Scission of the cyclohexadiene intermediate (VI) via pathway A leads to intermediate VII. Another chain scission followed by H/D abstraction from an adjacent chain leads to toluene (VIII) in which one hydrogen atom originates from an intermolecular reaction. Alternatively, pathway V leads to an intermediate (FX) containing a benzylic chloride. Chlorine replacement by H/D leads to intermediate X. Subsequent chain scission and H/D addition leads to toluene (XI) in which two hydrogen atoms originate from intermolecular reactions. Note in Figure 3 that the two key stable intermediates, VII and X, can "preform" in the polymer during the early stages of pyrolysis (i.e., during dehy drochlorination) . There is some experimental evidence for the formation of these types of "benzenoid" structures in partially degraded PVC. If one makes the assumption that each pathway has approximately equal probability, predictions can be made as to the expected isotopic abun dances of the toluene product. We will assume a primary kinetic isotope affect (k-H/kJD) of 2.0 which is reasonable for a high temperature reaction t~600C). The results of the predictions (Figure 4) show rather good agreement with the actual isotopic abundances (Table 1). The most abundant species are C7H8 andC^HD?. This agreement suggests that Figure 3 and the above assumptions may be quite reasonable. Also note that it was unnecessary to invoke crosslinking reactions (intermolecular C-C bond formation) as a route to transfer H/D. An alternative to Figure 3 would be the formation of intermediates in which methyl groups attached to cyclohexadiene rings are "preformed" in the earlier stages of polymer decomposition from the cyclohexadiene intermediate (VI). However, the kinetic isotope effect predictions (Figure 4) would be the same. Formation of Other Aromatic Pyrolyzates Two pathways are possible for styrene. One is completely intramole cular and requires no intermolecular H/D mixing. The other requires one intermolecular H/D transfer. Using the same type of analysis scheme as used for toluene results in a predicted isotope distribution in reason ably good agreement with the experimental values. The indene analysis parallels that of styrene; one intramolecular pathway and one requiring an intermolecular H/D transfer. Again, the agreement between experiment and prediction is reasonable. Naphthalene, like benzene, forms mainly from intramolecular pathways. A formation BFG05192 lOQZZLOZ 11 mainly related to changes in the environment around the molybdenum ion. Some comments about particle size and dispersion characteristics, how ever, are in order. Although M0O3 is the most effective compound, it and ammonium dimolybdate have the poorest dispersion characteristics. Grinding the ammonium dimolybdate to reduce its average particle size improved its performance, so it behaved much like the ammonium octamolyb- date. In contrast, while grinding improved the dispersion characteris tics of the M0O3, it did not improve its performance. In addition, commercially available samples' of molybdenum trioxide, designed to have higher surface areas and smaller particle sizes gave the same level of performance. ', -............ We decided to emphasize the development of molybdenum compounds as fire and smoke retarder ^additives.' Because many of them are quite effective, white,~ and in small scale laboratory studies process reason ably well into PVC. In particular, we decided to place our initial emphasis on molybdenum trioxide, M0O3, because it is the least expensive and most readily available molybdenum compound, and because it has a very favorable smoke reduction-concentration effect. Figure 9 illustrates that increasing the loading of M0O3 from 2 to 10 phr continually reduces smoke formation. _It must be mentioned again that such relationships only hold over a limited range of concentration. For example, in the case of M0O3, the concentration-loading curve normally plateaus out around 10 phr. So additional M0O3 above this level will provide relatively small improvements in performance. Other potentially interesting molyb denum compounds such as ammonium dimolybdate exhibited a very flat smoke-concentration relationship (Figure 10). . Smoke Reduction and Char Formation The relationship between smoke and char is critical in understanding the role of M0O3 and other metal smoke retarders in PVC. All of the effective smoke retarders we have studied effectively act to promote the formation of char while they decrease the formation of smoke. The smoke-char relationship is illustrated for a variety of metal smoke retarders in Tables 5 and 6. As is illustrated, M0O3 is one of the most effective char formers. Essentially all of the effective metal based smoke retarders studied appear to work in the solid state, not in the vapor state. While the possibility that some of the retarders provide a vapor phase activity cannot be ruled out, their main function appears to be that of interfering with the normal thermal degradation pattern of the PVC. The most visible manifestation of this solid state activity is the formation of char. All of the effective smoke retarders reported here promoted the formation of char. Char yields of 30-50% BC are typical. This compares to % BC of about 10 for the rigid control compound (Table 4). However, in a series of closely related smoke retarders or retarders used at different loading levels, there may be a non-regular relationship between smoke reduction and char formation. For example, in such a series, the compound giving the lowest smoke may not always form the most char. But using the Goodrich Smoke-Char test and the flaming mode of the NBS Smoke Chamber, even in these cases, smoke reduction is accompanied by char formation. BpG05194 20732009 13 cyclization to form benzene, are stable to higher temperatures where different mechanisms ensue to give aliphatic (less smoky) products. The Bell reports considered two possibilities for trans polyene formation. One is the straightforward Lewis acid isomerization of cis alkenes into trans alkenes. The other has the M0O3 acting as a Lewis acid which interacts with chain chlorines and induces dehydrochlorination. If a dissociated ion pair forms without retention of the original stereochemistry, trans polyenes can be formed directly during dehydro chlorination. We did not believe the "cis-trans1* Lewis acid mechanism as origin ally proposed by Bell Laboratories' workers9 explains the primary role of M0O3 as ;a smoke retarder-in PVC. It is weak in several respects, both on its own merits and also with regard to the results presented here and in previous work.8 Our most serious objection to the "cis-trans" Lewis acid theory is with regard to its predictions of volatile pyrolyzate formation. Lewis acids can promote the isomerization of both cis and trans double bonds; the trans configuration is thermodynamically more stable.- Even if trans polyenes are formed initially upon dehydrochlorination, we would expect rapid cis-trans interconversion to occur at the high temperatures and enthalpies encountered in later stages of PVC thermal decomposition. This might lead, for example, to enhanced evolution of benzene at higher temperatures, yet the evidence does not show this. The Lewis acid theory also predicts that aliphatic products will be formed from polyene chains in preference to aromatics. The theory predicts that these aliphatics will burn cleanly (with little smoke). These compounds, however, will in general be unsaturated since they are derived from polyene chains. Thus, they should burn with a smoky flame, similar to benzene and other aroma tics . In fairness to the Bell workers, in their detailed discussion of the Lewis acid mechanism, they do point out that in addition to cis-trans effects, M0O3 or species derived from it might destroy the olefinic precursors of benzene by catalyzing intermolecular Diels-Alder cycylizations or Friedel-Crafts alkylations. This would lead to crosslinking of the thermally decomposing PVC chains segments. However, they state that completely convincing correlations of char yield with smoke emission or Lewis acid content have not been observed in published combustion stu dies. And they conclude "that crosslinking catalyzed by metal species has not been established as the principle mechanism for smoke inhibition, although it undoubtedly occurs in many systems". Our mechanistic studies were undertaken with two thoughts in mind. One was to test the Lewis acid mechanism, the other was to search for alternate explanations consistent with our experimental results and the published results of other workers. Smoke and Char Formation The model PVC compounds (with and without M0O3) were examined in the flaming mode of the NBS smoke chamber, and also in the Goodrich Smoke- QTOZZLQZ BFG05195 15 tic PVC (SYN), (3) 100 parts PVC plus 10 parts M0O3, and (4) 100 parts SYN plus 10 parts M0O3. Pyrolyses were carried out in a helium atmos phere at 550C for 20 sec. Quantitative determinations for benzene and toluene pyrolyzates were made on each sample. Response factors were determined (external standard method) by injection of standard solutions of benzene and toluene (n-pentane solvent; porapak PSGC column). The syndiotactic PVC sample was prepared by a urea conrplexation method and was reported to be nearly 10Q% syndiotactic. Samples 3 and 4 were mixed with an agate mortar and pestle. The results are reported in Table 9. Instead of the expected 50% or better reduction in benzene relative to the controlled PVC sample, only a 25% reduction was observed. Also, if the principle action of the smoke retarder is to induce the formation of trans polyene segments on dehydrochlorination, one would expect the metal additive to have less effect in syndiotactic PVC than in ''normal" 103EPF76. In fact, M0O3 reduced benzene by only approximately 57% in 103EPF76 compared to approximately 72% in syndiotactic PVC. The M0O3 smoke retarder also is effective in reducing toluene formation, but the 40-50% reduction is smaller than for benzene. We conclude that M0O3 must be doing more than acting as a Lewis acid isomerization catalyst. Perdeutero-PVC Pyrolysis Results Two PVC samples were used. The "normal" PVC sample was Geon 103EPF76 prepared by suspension polymerization at 50C. (This was the same PVC sample used for the smoke, char, and volatile pyrolyzate tests.) The perdeutero-PVC samples (DPVC) was prepared by the suspension polymeriza tion of C2D3CI (reported_deuterium enrichment 97.3%) at 50C. GPCderived parameters were Mn = 1.9x10s and Mw = 3.9x10s. Deuterium enrichment experiments were conducted with the following simple compounds: (1) PVC/DPVC (coprecipitate) and (2) 100 parts PVC/ DPVC plus 10 parts M0O3. Pyrolyses were carried out in helium atmosphere at 550C for twenty seconds and deuterium enrichments lor selected pyrolyzates were calculated from the field ionization mass spectral data. In each experiment the polymer sample used was an equal weight coprecipi tate of PVC (103EP-F76) and DPVC. The PVC/DPVC coprecipitate and M0O3 were mixed with an agate mortar and pestle. About 2 mg of sample was pyrolyzed in each run; duplicate runs were made on both Dexsil 300 and Poropak PS-GC columns. The isotopic distributions for selected pyrolyzates are listed in Table 10. Corrections have been made for natural abundance of 13C. No corrections for molecular ion fragmentation were necessary, since pure molecular ion spectra were obtained by field ionization. All pyrolyzates listed, excepted for benzene, showed very extensive H/D mixing in the M0O3 compound. Benzene gave relatively little H/D mixing indicating that benzene must be formed intramolecularly even with the smoke retarder present. The total amount of benzene formed with the smoke retarder was, of course, greatly reduced (see Tables 8 and 9). With M0O3 present, naphthalene and the other aromatic pyrolyzates containing alkyl groups all showed considerable H/D mixing. Their formation must involve intermolecular reactions. Such intermolecular reactions could be crosslinking of the decomposing polymer chains follow- BFG05196 20732011 (a) Allylic Site Coupling 17 (b) Alkyl Site Coupling Reductive coupling reactions are quite common in organometallic chemi stry. Interestingly, the reaction sequences seem best established for copper compounds, the most effective class of smoke retarders for PVC. With copper as the metal, 1,1 elimination (or "cross-coupling") may be a feasible pathway to effect reductive coupling. This involves a single cation and can be formalized as follows: R- + Cu -- R--Cu1 RCu1 + R' - Cl --. (RR'CumCl] [RR'CuinCl] -- R--R' + Cu'Cl HH H HH (3) (4) (5) where Rjt' ---V--c---c--V-- <* --V--c--c=c--y-- H This pathway involves two oxidative additions, reactions 3 and 4, followed by the reductive elimination, reaction 5, tojjve the coupled product (R-R1). An alternate means of forming the Cu intermediate would be the oxidative addition of two radicals (R* and R') to Cu Cl. The initiating radical can be formed via PVC chain scission reactions or else via chlorine extraction by the metal additive.2'3 R - Cl + Cu -- R- + Cu*Cl (6) Although we felt less confident about postulating a "reductive coupling" scheme based on low valence molybdenum species, we proposed a mechanism based on a dinuclear transition state exhibiting Mo-Mo bonding. R - Cl + Mon -- [R - Mon+2 - Cl] -- Mort+1 - C1 + R- (7) 2[Cl - Mon+2 - R] Dinuclear Transition State(s) R - R + 2Mon+1 -- Cl (8) BFG05197 20732012 19 Our secondary criticism of the Bell model compound studies is the low temperatures used. Most of their reactions were carried out between 100 and 200C. These are unrealistically low thermal degradation temper atures for smoke retarded PVC. From about 17 experiments, only 2 were conducted at temperatures as high as 300C. Our final criticism of the Bell model compound studies concerns their failure to adequately report the details (techniques, equipment, procedures) of their experiments. They do not provide sufficient infor mation to enable other researchers to duplicate their experiments and check their results. 2,4,6-Trichloroheptane Model Compound .... Our model compound is a well-recognized model for PVC -- 2,4,6-tri chloroheptane: Cl Cl Cl In contrast to the model compounds used by the Bell workers, trichloroheptane can "unzip" during pyrolysis to form a polyene chain with three double bonds. This chain can then intramolecularly cyclize to form an aromatic ring. Crosslink formation between adjacent chains also can provide a pathway for forming condensed ring aromatic compounds. Pyrolysis The pyrolyses were conducted in glass capillary tubes which were purged with nitrogen and sealed. Smoke retarders were either dissolved in the 2,4,6-trichloroheptane prior' to charging, or uniformly coated on the walls of the capillary tubes. The synthesis of 2,4,6-trichlorohep tane used in this paper involved four steps starting with dehydroacetic acid. 8o(OH)2 000 OH OH OH The first step involves the acetic rearrangement of dehydroacetic acid. In the second step the 2,6-dimethyl-4-pyrone is ring-opened with barium hydroxide. The third step, hydrogenation, must be done soon after the second step because the 2,4,6-heptanetrione does not have a very long shelf life. It enolizes very easily and this reduces the yield in step 3, which uses a ruthenium catalyst. Step 4, the reaction with thionyl chloride, is a standard procedure for the conversion of alcohols to chlorides. The final product was distilled twice to separate it from undesired side-products from the chlorination reaction which had similar boiling points. 20732013 BFG05198 21 SCHEME I MW 190 Dimer (several isomers possible) Dihydro-oligomers were also detected in the FD-MS analysis of the decomposition products. The field desorption results are listed in Table 14. Since copper compounds are among the most effective catalysts for promoting reductive coupling type reactions, it is reasonable to suggest that the dihydro-oligomers formed via this pathway. The Molybdenum Trioxide Experiment This was the most interesting experiment since it gave unexpected results. The first surpise was the formation of a considerable quantity of low-boiling components. Significant low-boils were not formed in the control and copper catalyzed experiments. Headspace analysis, in which the pyrolysis tube was placed in a sealed sample file and subsequently broken, enabled the low-boils to be sampled and analyzed by GC-MS. The following low-boiling components were detected: A) .Ci-Cy saturated (or alicyclic) hydrocarbons, either straight chain or methyl-branched. B) C1-C7 chlorinated, saturated hydrocarbons. C) Toluene. GC-MS and direct probe MS were used to identify the higher-boiling components. The capillary gas chromatogram of the high-boils was very "messy" and contained a large number of isomers for most of the observed species. A listing of the most prominent components is provided in Table 15. The dominant products were"alkyl-substituted benzenes and naphthalenes, although some of the normal dimer and its dihydro deriva tive also were observed. In addition to the products listed in Table 15, there were many more products of higher molecular weight; these were detected using direct probe electron impact MS. The results of the direct probe analysis are summarized in Table 16. The trend established in the results shown in Table 15 continues. Clearly a primary role of BFG05199 23 Experimental assistance from Edgar R. Harris, Umars Sockis, and Betty A. Starkey is gratefully acknowledged. Helpful support and suggestions were provided by E. Douglas Dickens, Jr., Arthur W. McRowe, and Paul P. Nicholas. REFERENCES 1. R. L. Rawls, Chem. Eng.'News, 6_1 (1), 9 (1983). 2. "' R. P. Lattimer and W. J. Kroenke, J. Appl. Polm. Sci., 25, 101 (1980). - 3. W. H. Starnes, Jr. and D. Edelson, Macromolecules, 12, 797 (1979). 4. T. Kelen, J. Macromol. Sci. Chem., A12, 349 (1978). 5. B. B. Troitskii, L. S. Troitskaya, V. N. Myakov, and A. F. Lepaev, J. Poly. Sci., Polym. Symp., 42, 1347 (1973). 6. S. A. Liebman, J. F. Reuwer, Jr., K. A. Gollatz, and C. D. Nauman, J. Polym. Sci. Part A-l, 9, 1823 (1971). 7. W. H. Starnes, Jr., L. D. Wescott, Jr., W. D. Reents, R. E. Cais, G. M. Villacorta, I. M. Plitz, and L. J. Anthony, Org. Coatings Plast. Chem., 46, 556 (1982). 8. W. J. Kroenke, J. Appl. Polym. Sci., 26, 1167 (1980). 9. D. Edelson, V. J. Kuck, R. M. Lum, E. Scalco, W. H. Starnes, Jr., and S. Kaufman, Combust. Flame, 38, 271 (1980). 10. R. P. Lattimer and W. J. Kroenke, J. Appl. Polym. Sci., 26, 1191 (1981). 11. D. Edelson, R. M. Lum, W. D. Reents, Jr., W. H. Starnes, Jr., and L. D. Wescott, Jr., "New Insights into the Flame-Retardance Chemi stry of Poly(vinyl chloride)", presented at the 19th Inti. Symp. on Combustion, Haifa, Israel, 1982. s t o z e z .02: BFG05200 25 FIGURE 2 BENZENE FORMATION s'. X BfG05T01 20732016 Li FIGURE 4 PREDICTED ISOTOPIC ABUNDANCES for TOLUENE PATHWAY A, SCHEME H ,0.67n CtH8 y nCrH7 (from PVC1 +0\ tiihQ nC707 (from OPVC1 ,0.670 070-^ PATHWAY 8, SCHEME H ,, u_fl,.y0-44nC7,H8 0.67nR|C7H7 Am_ +H/ '*Nl22nC7H70 nR|C7H5 (from PVQ +0\ _fl .a22nC7H70 '*`1ArH Q.33rR,C7H60 0.1 In C7H5D2 +H, nR|C70g'" (from 0PVC1 +0 _fl^/a44nC70eH2 ,Q.67nR|C706H _ +Sl22n C707H ^ 0.22T1C707H >a33nR,C707^ +\o-X*0i.11 nC70g SUMMATION of Ihe VARIOUS PATHWAYS! C7He C7H7Q C7Hg02 C7H503 C7H4D4 Cr^Os 07^2^6 C7HD7 C7O3 TOTAL 1.1 In 0.78 n Ctl In 0 ' 0 0 0.44n l.l In 0.44n PERCENT 28 19 3 0 0 0 11 28 II 20732017 BFG05202 29 FIGURE 6 // Group I-B Transition Metals Cu(AcAc)2 I----I Cul M Cuo H Cu20 | 1 Cuso4 hH 70 60 50 40 30 20 20732018 BG05203 31 FIGURE 8 Group 21-B Transition Metals // I --I |----| Cr (AcAc)3 I---------- *---------- 1 CrCI3 (NH4)MOg02g || (NH4)2Mo207 I-- ----1 Na2Mo04 J--#H CaMo04 | | SrMo04 |1 ZnMo04 H Mo03 [ j h2wo4 |-------------- W03 |------------- 1 70 60 50 40 30 20 V9 B?G5a4 20732019 FIGURE 10 BFG05205 OZOZSJ-OZ 35 : FIGURE 12 v 0526 20732021 37 FIGURE 14 BFG05207 20732022. 39 TABLE 1 Isotopic Distributions of Selected PVC/DPVC Pyrolyzates Isotopic species C|HmDn*MW Observed* Percent abundance Corrected6 Benzene CH-78 CHD-79 CH*Dt-80 ' C4H3D3-8I CeHsD4-82 CHI>-83 C$Da-84 Toluene CtH-92 CtH7D-93 CTH^Df94 C7HsDj-95 C7H4D4-96 CtHjDs-97 C7H,D-98 C7HD7-99 C7D8-IOO Styrene CsHs-104 CsH7D-105 CHD7-106 CaH|D3*107 CsHJ34-108 CsH3Ds-109 CgHjDs-110 CHD7-111 CsDg-112 Indene CsHg-116 CH7D-U7 C9HD2-118 CeHsDs-119 C9H434-120 C9H3D5-I21 CH,D-122 CoHDt*123 CsDg-124 Naphthalene CioHa-128 CiqHt0*129 CioHeDi-130 C10H5D3-I3I CioH4D4-132 C10H3DS-I33 - C|oH]Da*134 CloHDt-135 - CioOa-136 l- and 2-Methylnaphthalene CuHio*142 C11H9D-143 CiiHgDt-144 C11H7D3-I45 ' SO 2.6 U 0.6 1.4 7.5 36 23 17 8.2 1.7 2.7 4.8 13 22 7.4 37 10 4.1 2.5 0.8 0.7 6.7 16 22 42 8.3 4.9 1.9 1.6 4.0 6.0 11 21 57 5.1 0.9 0.7 0.1 0.8 1.0 6.3 28 25 18 15 4 50 3 2 <1 l 2 42 23 17 8 2 3 4 10 24 9 37 10 4 2 1 1 4 14 27 42 8 5 2 2 3 5 7 26 57 5 1 1 ~0 1 1 l - 34 25 18 15 4 Predicted* 50 0 0 0 0 0 50 28 19 3 0 0 0 11 28 11 42 8 0 0 0 0 0 17 33 42 8 0 0 0 0 0 17 33 67 0 0 0 0 0 0 0 33 37 26 4 0 20732023 BFG05208 41 TABLE 3 Model Rigid PVC Compound: Smoke Formation in the NBS Smoke Chamber Test Tin stabilizer An/g an no 57.6 4.2 6 yes 64.9 9.3 - - 147 ; TABLE 4 Model Rigid PVC Compound: Smoke and Char Formation in the Smoke-Char Test Tin stabilizer 5pvc a n %BC a n no 103.0 16.9 9 11.0 6.6 9 yes 101.8 14.4 130 8.0 4.2 53 0732024 BG0W9 *5 TABLE 6 SMOKE-CHAR CORRELATION - M0O3 M0O3 (phr) - 2 5 10 Dm/g 55 40 28 21 spvc 116 48 43 40 % BC 9 28 35 37 20732025 b^G05^ 45 TABLE 8 A. Relative Abundances of Selected Pyrolyzates at 550C Pyrolyzate Relative abundance* C i--C;i aliphatics -- C^-Cfi aliphatics Methyl and ethyl chlorides Benzene ." Toluene Dimethyl- and ethylbenzene Naphthalene Methylnapthalenes Biphenyl :- 2.3 1.5 2.9 0.20 0.91 0.78 0.14 0.35 0.37 . Total aliphaticsb Total aromatics' Total chlorod 2.1 0.60 B. Computer Area per mg Sample (Arbitrary Units) Total aiiphaticsb Total . aromatics' Total chlorod Control MoO:i-PVC 6,900 14,200 25,600 15,400 14 42 a Ratio of FID-GC peak area of MoO;i-PVC compound to the control compound (average of multiple runs). b Sum of Ci-Ce aliphatic hydrocarbons. c Sum of all aromatic compounds detected. d Mostly methyl chloride and ethyl chloride. 20732026 47 TABLE 10 dx-MW1' do-28 d, -29 da-30 d;,-:u d.,-32 Isotopic Distribution of Selected PVC Pyrolvzates1'' Control' MoOad dx-MW Control C^H4 Ethylene 12. 16. :)i. 27. 14. /s ;i.6 22. 62. 24. 19. d(l-92 d,-9:i da-94 da-96 d4-96 dr,-97 d-98 d7-99 ds-KH) CtHh Toluene 6.1 14. 15. 11. 11. III. 16. 11. :i.;i MoO;, 5.7 8.6 13. 14. 17. 16. 13. 10. 4.0 d,,-78 d,-79 d>-80 d:t-81 d4-82 d:-.-83 du-84 CnH,; Benzene 5 2. 4.7 l.;l 0.6 1.9 8.1 31. 4.!. 10. 3.7 ;i.6 4.6 9.4 26. do-116 d, -117 da-118 da-119 d4-l20 dr,-121 do-122 d7-123 da-124 C.iHs Indene 28. 11. 9.6 9.2 6.9 8.:? 11. 11. 10. 0. 3.3 8.5 15. 17. 24. 15. 12. 5.3 C7toHtt Naphthalene CuHio Methylnaphthalene do-128 d, -129 da-130 d:1-i:u d4-132 cU-i:u do-184 d7-135 dM-136 45. 8.7 4.4 2.2 1.4 1.6 ;).5 11. 22. 28. 10. 10. . 7.5 7.1 9.9 3.7 9.5 14. do-142 d,-148 da-144 da-146 d4-146 dr,-147 d-14S d--149 d-l50 dg-161 dio-152 10. 15. 14. 9.6 7.0 5.4 6.7 8.9 11. S.:i 3.0 2.3 4.3 8.7 9.4 17. 18. 15. 11. 11. 2.5 1.7 " Determined from PY-GC-FI-MS molecular ion intensities; average of duplicate runs. abundances have been corrected for the natural abundance of l:'C. 11 Isotopic species (dT) and molecular weight (MW). '' Percent isotopic abundances for the "control" sample are repeated from ref. 11. H Control plus 10 phr MoOn. Isotopic 20732027 BFG05212 f. TABLE IB Higher-Boiling Products from the Pyrolysis of 2,4,6-Trichloroheptane _ in the Presence of a Molybdenum Catalyst as Determined by (GC)2-MS Molecular Weight Compound Description 92 CH34> Toluene 106 c2<t> 3 isomers 120 C3<t> 4 isomers - mostly methyl-ethyl 134 c4<t> 4 isomers - many are dimethyl-ethyl and/or methyl-isopropyl 148 Cs<f . 7.isomers - some are methyl-diethyl , and/or methyl-isobutyl 162 c6<C 2 isomers 170 C3Np Substituted naphthalenes-5 isomers mostly methyl-ethyl and/or isopropyl 184 C4Np 6 isomers - mostly methyl-isopropyl .. and/or dimethyl-ethyl and/or diethyl 188 Dimer 2x (I) 190 Dimer I + 2H 198 CsNp 7 isomers 20732029 BFG05214